Pika

Pika Video Generation API (via fal)

Pika's official developer access is not hosted by Pika. Pika partnered with fal to serve its Pika 2.2 video models on fal's inference infrastructure, so the pika.art/api page hands developers off to fal.ai. On fal, Pika models are exposed as async queue endpoints (submit / status / result) authenticated with a fal API key (Authorization: Key FAL_KEY). Documented model endpoint IDs include fal-ai/pika/v2.2/text-to-video, fal-ai/pika/v2.2/image-to-video, Pikascenes, and Pikaframes, plus v2.1 variants. Inputs include prompt, resolution (720p/1080p), aspect_ratio, duration (5 or 10 seconds), seed, and negative_prompt. Endpoints below are modeled from fal's published model pages, not from a Pika-operated OpenAPI definition. No first-party Pika API, OpenAPI document, or public WebSocket surface is published on pika.art.

Documentation GitHub

Documentation

https://fal.ai/models/fal-ai/pika/v2.2/text-to-video/api

Other Resources

🔗

Announcement

https://blog.fal.ai/pika-api-is-now-powered-by-fal/

🔗

APIsJSON

https://raw.githubusercontent.com/api-evangelist/pika-art/refs/heads/main/apis.yml

API entry from apis.yml

aid: pika-art:pika-art-video-generation-api
name: Pika Video Generation API (via fal)
tags:
- Video Generation
- AI Video
- Text-to-Video
- Image-to-Video
image: https://kinlane-images.s3.amazonaws.com/shared/apis-json/apis-json-logo.jpg
humanURL: https://pika.art/api
baseURL: https://queue.fal.run
properties:
- url: https://pika.art/api
  type: Documentation
- url: https://blog.fal.ai/pika-api-is-now-powered-by-fal/
  type: Announcement
- url: https://fal.ai/models/fal-ai/pika/v2.2/text-to-video/api
  type: APIReference
description: 'Pika''s official developer access is not hosted by Pika. Pika partnered with fal to serve
  its Pika 2.2 video models on fal''s inference infrastructure, so the pika.art/api page hands developers
  off to fal.ai. On fal, Pika models are exposed as async queue endpoints (submit / status / result) authenticated
  with a fal API key (Authorization: Key FAL_KEY). Documented model endpoint IDs include fal-ai/pika/v2.2/text-to-video,
  fal-ai/pika/v2.2/image-to-video, Pikascenes, and Pikaframes, plus v2.1 variants. Inputs include prompt,
  resolution (720p/1080p), aspect_ratio, duration (5 or 10 seconds), seed, and negative_prompt. Endpoints
  below are modeled from fal''s published model pages, not from a Pika-operated OpenAPI definition. No
  first-party Pika API, OpenAPI document, or public WebSocket surface is published on pika.art.'